gke/ERR/2021_006

GKE Autoscaler isn’t reporting scaleup failures.

Product: Google Kubernetes Engine
Rule class: ERR - Something that is very likely to be wrong

Description

If the GKE autoscaler reported a problem when trying to add nodes to a cluster, it could mean that you don’t have enough resources to accommodate for new nodes. E.g. you might not have enough free IP addresses in the GKE cluster network.

Remediation

You can use the following filter to find matching log lines:

resource.type="gce_instance"
log_id(cloudaudit.googleapis.com/activity)
severity=ERROR
protoPayload.methodName="v1.compute.instances.insert"
protoPayload.requestMetadata.callerSuppliedUserAgent="GCE Managed Instance Group for GKE"

Further information